AI Fluency Program
Three sessions. Real workflows. Teams leaving with more than just exposure.
A structured cohort program that takes teams from a mix of AI users and avoiders to a shared baseline of fluency. Grounded in the behavior change methodology I refined over a decade in performance science consulting, applied to the specific challenge of building AI fluency at the team level.
Each participant leaves with at least one AI workflow they can use on Monday morning, and the team leaves with a shared vocabulary and a way of thinking about AI together.
The Flow
- Pre-work: light async setup so session one starts at speed.
- Session 1 — Demonstration: see what's possible, calibrate expectations.
- Session 2 — Practice: hands-on workflow design with the actual work people do.
- Session 3 — Design & Sustain: lock in workflows and the practice that keeps them growing.
- Deliverables: workflow library, fluency baseline, sustained practice plan.
The AI Fluency Ladder
| Level | Archetype | Behavior |
|---|---|---|
| 0 | Non-User | Haven't meaningfully used AI for work |
| 1 | Information Seeker | Treating AI like Google |
| 2 | Task Delegator | Using AI as an assistant for discrete tasks |
| 3 | Thinking Partner | Bringing half-formed ideas and iterating to better answers |
| 4 | Workflow Designer | Building repeatable processes with templates and prompt sequences |
| 5 | Artifact Builder | Creating durable tools (scripts, websites, apps) |
| 6 | System Orchestrator | Composing agentic systems that take multi-step actions |
The Principles
The left side names the failure mode to avoid. The right side names the behavior to adopt instead, and the human skill it parallels.
| # | Avoid: pitfall | Adopt: principle | Result: human-skill parallel |
|---|---|---|---|
| 1 | I don't know what to ask | Have AI interview you first | Coach by asking, not telling |
| 2 | I jumped to the deliverable | Start with the goal, not the output | Align on outcomes before activities |
| 3 | AI agreed with my hypothesis | Don't lead the witness | Get unbiased input before sharing your view |
| 4 | I summarized instead of sharing the source | Bring primary sources, not summaries | Share raw context with collaborators, not pre-chewed conclusions |
| 5 | One prompt, generic answer | Sequence the work; don't cram one prompt | Design processes, don't just execute tasks |
| 6 | AI flattered me | Tell AI to push back | Build cultures where people push back |
| 7 | I shipped the first draft | Use AI to critique, not just create | Invite challenge to your work before shipping |
| 8 | I quit after one response | Iterate; the good answer is on round 4 | Refinement is where quality lives |
| 9 | AI gave generic output | Assign a role before asking | Clarify which hat someone is wearing |
Pre-work
- Short async survey to assess each participant's starting fluency level.
- Identification of one real piece of work each participant will redesign during the program.
- Light reading and tool setup so session one can be hands-on from the first minute.
Session 1 — Demonstration
- Live demonstrations of AI workflows that solve real problems the team faces.
- Calibration: what AI can and can't do well right now.
- Outcome: shared baseline of "what good looks like" for this team.
Session 2 — Practice
- Each participant builds at least one AI workflow from their actual work.
- Live coaching, peer review, iteration.
- Outcome: every participant has a working AI workflow they will use after the session.
Session 3 — Design & Sustain
- Workflows are documented, shared, and integrated into the team's operating cadence.
- Practice plan defined: how the team will keep fluency growing month over month.
- Outcome: a workflow library, a fluency baseline, and a practice plan that holds.